Radiopharmaceutical therapy (RPT) is one of the few viable alternatives to chemotherapy for patients with metastatic cancer. In non-Hodgkin's lymphoma (NHL), RPT has yielded durable remissions in treatment- refractory patients. RPT is the only treatment for neuroendocrine and metastatic thyroid cancer and is an emerging treatment for metastatic ovarian, prostate and other cancers. ClinTrials.gov lists 136 radionuclide therapy trials. RPT is administered like chemotherapy, by assuming that a maximum tolerated administered activity (AA) defined in a dose escalation trial applies to all patients. Such generic dosing has led to conservative treatment, yielding low toxicity at the expense of tumor control. With prior NIH support we have developed a patient-specific dosimetry (PSD) methodology and have shown it to be superior to generic treatment by enabling, for example, more aggressive yet safe therapy of diffuse lung metastases in thyroid cancer and a combined XRT/RPT treatment plan for osteogenic sarcoma, boosting tumor dose while keeping adjacent spinal cord dose below the MTD. The objectives of this competing renewal application are to further improve accuracy and to evaluate overall impact on RPT. Specifically: 1. We propose to develop a method to enable micro-scale dosimetry from macro-scale (imaging) data. Imaging-based PSD accuracy is limited by imaging resolution. In some cases, micro-scale absorbed dose (AD) distributions are key to understanding and thereby avoiding normal organ toxicity. 2. In evaluating impact, statistical uncertainty is important to interpreting results and guiding treatment. We will develop a method to calculate the uncertainty and confidence level of dosimetry results. 3. Accrual of a large number of dose-response studies, in a standardized manner, is needed to evaluate the impact of Aims 1 and 2, and PSD, generally, on improving tumor control with RPT. The software package, 3-D Radiobiological Dosimetry (3D-RD) developed with prior NIH support, and revised in Aims 1 and 2, will be used to perform PSD calculations for a large number of existing and prospective, in- house, and collaborating institution studies. Single-institution studies yield limited data; 3D-RD analysis for collaborator studies leverages data from other sites and increases the patient population pool to yield a robust data set for dose-response studies. 4. 3D-RD includes radiobiological modeling for dose rate and dose non- uniformity. Parameters values for these models cannot currently be measured in individuals. Instead, literature values are used. Without standardization, different investigators/institutions will use different values making response comparisons across studies difficult. To support the standardization needed for the dose-response studies of aim 3 we will establish an on-line database of reference radiobiological parameter values. Such a database would be analogous to the ICRP reference man compilation of organ masses and compositions. RPT is a promising treatment for metastatic cancer. RPT is currently delivered according to a chemotherapy paradigm. Support for this proposal will help bring a rational, AD-based approach, to RPT delivery.